linux-kernel - Re: PCIe/PCI rescan for new FPGA devices

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Thu, 28 Apr 2011 23:40:58 -0700
From:	Yinghai Lu <yinghai@...nel.org>
To:	Xianghua Xiao <xiaoxianghua@...il.com>
Cc:	linux-kernel@...r.kernel.org, linux-pci@...r.kernel.org
Subject: Re: PCIe/PCI rescan for new FPGA devices

On Thu, Apr 28, 2011 at 4:06 PM, Xianghua Xiao <xiaoxianghua@...il.com> wrote:
> I have a FPGA-PCIe device and a FPGA-PCI device that are empty at
> kernel boot time, they're only loaded after kernel is up. After I load
> FPGA images, "lspci" can not show the device, "echo 1 >
> /sys/bus/pci/rescan" also does nothing about it. With a reboot I can
> see the devices are allocated correctly though, but I want to avoid
> the reboot.
>
> I checked at fakephp(which is to be deprecated) and hotplug/rescan,
> none of that worked for this scenario.
>
> Can 'rescan' handle cases like this? or, is there a way that I tell
> the kernel to reserve a few BARs somehow in the PCI topology at
> bootime so I can "insert" the FPGAs later?

well, you did not send out the lspci -tv yet.

assume lspci tree like this

 +-[0000:c0]-+-00.0-
 |           |
 |           +-01.0-[c3]--
 |           +-03.0-[c4-c9]----00.0-[c5-c7]--+-02.0-[c6]--+-00.0 your FPGA

so c0:03.0 is pcie root port,
c4:00.0 is your pcie switch upstream port
c5:02.0 is your pcie switch downstream port
c6:0 is your FPGA controller.

so you could remove c0:03.0 or c4:00.0 or c5:02.0, then do rescan.
--- depend if your BIOS allocate the right value to the c0:03.0

Yinghai
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/